A memory controller for controlling a memory operation of a memory device includes: an error correction code (ECC) circuit configured to detect an error of first read data read from the memory device and correct the error; an error type detection logic configured to write first write data to the memory device, compare second read data with the first write data, detect an error bit of the second read data based on a result of the comparing, and output information about an error type identified by the error bit; and a data patterning logic configured to change a bit pattern of input data to reduce an error of the second read data based on the information about the error type.
|
14. A method of controlling a memory device, the method comprising:
reading first read data from the memory device;
correcting bits of a first memory area in which an error has occurred in the first read data and inverting bits of a second memory area in which no error has occurred;
identifying a first error type in the first memory area and a second error type, which is different from the first error type, in the second memory area; and
writing second write data having a minimum error to the memory device, based on at least one of the first error type and the second error type.
1. A memory controller for controlling a memory operation of a memory device, the memory controller comprising:
an error correction code (ECC) circuit configured to detect an error of first read data read from the memory device and correct the error;
an error type detection logic configured to write first write data to the memory device, compare second read data with the first write data, detect an error bit of the second read data based on a result of the comparing, and output information about an error type identified by the error bit; and
a data patterning logic configured to change a bit pattern of input data to reduce an error of the second read data based on the information about the error type, wherein
the first write data is data in which the error of the first read data is corrected, and the second read data is data obtained reading the first write data written to the memory device.
18. A memory system comprising:
a memory device including a plurality of memory areas, each of the memory areas including at least one memory cell; and
a memory controller configured to control a memory operation of the memory device, wherein
the memory controller is further configured to:
read first read data stored in a first memory area and a second memory area from the memory device,
perform a correcting operation on data of the first memory area in which an error has occurred in the first read data,
perform an inverting operation on data of the second memory area in which no error has occurred in the first read data,
generate first write data based on the correcting operation and the inverting operation,
perform a comparing operation of comparing the first write data with second write data obtained by reading the first write data, and
identify a plurality of error types based on the comparing operation.
2. The memory controller of
the error type comprises a plurality of error types including a temporary error and a stuck error, and
the error type detection logic is further configured to identify at least some of the second read data as the temporary error and identify another some of the second read data as the stuck error.
3. The memory controller of
4. The memory controller of
the error type detection logic is further configured to receive error bit information from the ECC circuit and generate the first write data based on the error bit information, and
the error bit information comprises information corresponding to a bit at which an error has occurred in the first read data.
5. The memory controller of
compare, on a cell-by-cell basis, the second read data stored in cells of a first memory area with the first write data stored in the cells of the first memory area,
identify, as the stuck error, a bit of the second read data in a first cell when the bit of the second read data in the first cell and a bit of the first write data in the first cell are different from each other, and
identify, as the temporary error, a bit of the second read data in a second cell when the bit of the second read data in the second cell and a bit of the first write data in the second cell are the same as each other.
6. The memory controller of
compare, on a cell-by-cell basis, the second read data stored in cells of a second memory area with the first write data stored in the cells of the second memory area,
identify, as a hidden stuck error, a bit of the second read data in a third cell when the bit of the second read data in the third cell and a bit of the first write data in the third cell are different from each other, and
identify, as a normal bit, a bit of the second read data in a fourth cell when the bit of the second read data in the fourth cell and a bit of the first write data in the fourth cell are the same as each other.
7. The memory controller of
8. The memory controller of
9. The memory controller of
10. The memory controller of
the data patterning logic comprises a randomizer and a random seed generator outputting a plurality of random seeds, and
the randomizer is configured to randomize the input data, received from a host, based on the plurality of random seeds and output a plurality of pieces of random data.
11. The memory controller of
the data patterning logic further comprises a minimum error selector, and
the minimum error selector is configured to select, as selected random data, random data having a smallest error bit from among the pieces of random data based on the information about the error type and output the selected random data as second write data.
12. The memory controller of
13. The memory controller of
15. The method of
16. The method of
17. The method of
reading second read data, wherein
the identifying comprises comparing a result of the correcting and inverting with the second read data and identifying at least one of the first error type and the second error type according to a result of the comparing.
19. The memory system of
a first partial comparing operation of comparing the first write data stored in the first memory area with second read data stored in the first memory area, and
a second partial comparing operation of comparing data of the second memory area of the first write data with data of the second memory area of the second read data.
20. The memory system of
|
This application claims the benefit of Korean Patent Application No. 10-2019-0049828, filed on Apr. 29, 2019, in the Korean Intellectual Property Office, the disclosure of which is incorporated herein in its entirety by reference.
The disclosure relates to a memory controller, and more particularly, to a memory controller for identifying errors in data written to memory cells and avoiding identified errors, a memory system including the memory controller, and a method of operating the memory controller.
As non-volatile memory devices, resistive memories such as phase change random access memory (PRAM), nano-floating gate memory (NFGM), polymer RAM (PoRAM), magnetic RAM (MRAM), ferroelectric RAM (FeRAM), resistive RAM (RRAM), etc. are well known in the art in addition to flash memories. A resistive memory has a non-volatile characteristic of a flash memory, as well as a high-speed property of dynamic RAM (DRAM).
When a memory cell of a non-volatile memory device wears out, a data bit may be stuck to 0 or 1. In this case, in response to a read request of a host, a data bit having a value different from that of the data bit that has been previously written may be read out. For example, even if 0 is written to a memory cell in which a data bit is stuck to 1, 1 may be read out in response to the read request.
The disclosure provides a memory controller for determining the type of error in data written to a memory cell and avoiding a predetermined error, a memory system including the memory controller, and a method of operating the memory controller.
According to an aspect of the disclosure, there is provided a memory controller for controlling a memory operation of a memory device, the memory controller including: an error correction code (ECC) circuit configured to detect an error of first read data read from the memory device and correct the error; an error type detection logic configured to write first write data to the memory device, compare second read data with the first write data, detect an error bit of the second read data based on a result of the comparing, and output information about an error type identified by the error bit; and a data patterning logic configured to change a bit pattern of input data to reduce an error of the second read data based on the information about the error type, wherein the first write data is data in which the error of the first read data is corrected, and the second read data is data obtained reading the first write data written to the memory device.
According to another aspect of the disclosure, there is provided a method of operating a memory controller for controlling a memory device, the method including: reading first read data from the memory device; correcting bits of a first memory area in which an error has occurred in the first read data and inverting bits of a second memory area in which no error has occurred; identifying a first error type in the first memory area and a second error type, which is different from the first error type, in the second memory area; and writing second write data having a minimum error to the memory device, based on at least one of the first error type and the second error type.
According to another aspect of the disclosure, there is provided a memory system including: a memory device including a plurality of memory areas, each of the memory areas including at least one memory cell; and a memory controller configured to control a memory operation of the memory device, wherein the memory controller is configured to: read first read data stored in a first memory area and a second memory area from the memory device, perform a correcting operation on data of the first memory area in which an error has occurred in the first read data, perform an inverting operation on data of the second memory area in which no error has occurred in the first read data, generate first write data based on the correcting operation and the inverting operation, perform a comparing operation of comparing the first write data with second write data obtained by reading the first write data, and identify a plurality of error types based on the comparing operation.
According to another aspect of the disclosure, there is provided a memory system having a memory device and a memory controller. The memory controller: identifies a first memory constituent within the memory device having a fault, modifies write data, received from a host device, to generate modified data based on the first memory constituent and the fault, and stores the modified data in the memory device such that a first value within the modified data is stored in the first memory constituent. The memory controller modifies the write data to increase a likelihood that the first value matches a value produced by the fault.
Embodiments of the disclosure will be more clearly understood from the following detailed description taken in conjunction with the accompanying drawings in which:
Hereinafter, embodiments of the disclosure will be described in detail with reference to the accompanying drawings.
Referring to
According to an embodiment of the disclosure, the memory device 200 may include a plurality of memory chips and each of the first to Nth memory areas 210_1 to 210_N may correspond to a memory chip. Each memory chip may include a memory cell array for storing data and a peripheral circuit for performing memory operations, for example, a write/read circuit and a control logic.
Each of the first to Nth memory areas 210_1 to 210_N may include a plurality of memory cells respectively arranged in areas where a plurality of first signal lines intersect with a plurality of second signal lines. In an embodiment, the plurality of first signal lines may be a plurality of bit lines and the plurality of second signal lines may be a plurality of word lines. Each of the plurality of memory cells may be a single level cell (SLC) storing one bit or may be a multi level cell (MLC) capable of storing data of two bits or more.
When data of one bit is written to one memory cell, the memory cells may have two resistance distributions according to the written data. Alternatively, when data of two bits is written to one memory cell, the memory cells may have four resistance distributions according to the written data. In another embodiment, in the case of a triple level cell (TLC) in which data of three bits is stored in one memory cell, the memory cells may have eight resistance distributions according to the written data. However, the disclosure is not limited thereto, and in other embodiments, the memory cells may include memory cells capable of storing data of four bits or more, respectively.
Each of the first to Nth memory areas 210_1 to 210_N may include resistive memory cells each including a variable resistive element. For example, when the variable resistance element is a phase change material (e.g., Ge—Sb—Te (GST)) and the resistance of the variable resistance element varies with temperature, the resistive memory device may be phase change random access memory (PRAM). As another example, when the variable resistive element includes an upper electrode, a lower electrode, and a complex metal oxide therebetween, the resistive memory device may be resistive RAM (RRAM). As another example, when the variable resistance element includes the upper electrode of a magnetic body, the lower electrode of the magnetic body, and a dielectric therebetween, the resistive memory device may be magnetic RAM (MRAM).
According to an embodiment of the disclosure, the memory device 200 may be referred to as a resistive memory device as the memory device 200 includes resistive memory cells. Alternatively, in an embodiment of the disclosure, the memory device 200 may include various types of memory cells and as the memory cells are arranged in areas where a plurality of first signal lines intersect with a plurality of second signal lines, the memory device 200 may be referred to as a cross-point memory device. It is assumed, in the following embodiments, that the memory device 200 is a resistive memory device, but embodiments of the disclosure are applicable to various types of memory devices in which data of a memory cell may be stuck to a certain value. As an example, the embodiments of the disclosure may be applied to various types of memory devices such as a volatile memory device and a flash memory device.
In various implementations, the memory device 200 may be a memory module in which a plurality of memory chips are mounted on a module board. Alternatively, the memory device 200 may be a semiconductor package including a plurality of memory chips and the memory chips may be stacked, in a die form, in the semiconductor package, and each of the first to Nth memory areas 210_1 to 210_N may be a memory chip included in the semiconductor package.
Alternatively, the memory device 200 may be one semiconductor chip including a plurality of semiconductor layers. In this case, each of the first to Nth memory areas 210_1 to 210_N may be a semiconductor layer. Alternatively, when the memory device 200 includes one semiconductor layer and the semiconductor layer includes a memory cell array including a plurality of cell blocks, each of the first to Nth memory areas 210_1 to 210_N may correspond to a cell block.
The memory controller 100 may include a processor 110, an error correction code (ECC) circuit 120, an error type detection logic 130, and a data patterning logic 140, and the memory controller 100 may control memory operations in a manner that is based on hardware, software, and a combination thereof. Hereinafter, the error type detection logic 130 may be referred to as a stuck bit detection logic.
The memory controller 100 may control the memory device 200 to read data stored in the memory device 200 or to write data to the memory device 200 in response to a write/read request from the host. Specifically, the memory controller 100 may control program (or a write), read, and erase operations for the memory device 200 by providing an address ADD, a command CMD, and a control signal CTRL to the memory device 200. In addition, data DTA to be written to the memory device 200 and data DTA read from the memory device 200 may be transmitted and received between the memory controller 100 and the memory device 200.
According to various embodiments, the memory controller 100 may include RAM, a host interface, and a memory interface. The RAM may be used as an operating memory of the processor 110. The processor 110 may control overall operations of the memory controller 100. The host interface may include a protocol for performing data exchange between the host and the memory controller 100. For example, the memory controller 100 may be configured to communicate with the host via at least one of various interface protocols, such as universal serial bus (USB), multi-media card (MMC), peripheral component interconnect-express (PCI-E), advanced technology attachment (ATA), serial-ATA, parallel-ATA, small computer system interface (SCSI), enhanced small device interface (ESDI), and integrated drive electronics (IDE).
The memory controller 100 may read first read data from the memory device 200. The ECC circuit 120 may perform error detection and error correction on the first read data. In addition, the ECC circuit 120 may transmit information about error-corrected data and a data bit with an error to the error type detection logic 130.
The error type detection logic 130 may output first write data, obtained by inverting some data bits included in received data, to the memory device 200 to write the first write data to the memory device 200. In addition, the error type detection logic 130 may read the first write data written to identify an error type. In this case, the read first write data is referred to as second read data. Data bits of the first write data and data bits of the second read data may be different from each other. For example, when some memory cells are worn out and data is in a stuck state, a stuck error may occur in some data bits of the first write data. That is, the stuck error may be an error that outputs the same value regardless of the value of data written to a memory cell due to a physical or structural change of the memory cell. For example, a data bit of 1 is written, but a data bit of 0 may be read. Therefore, by comparing the first write data with the second read data, a data bit having an error and a memory cell that caused the error may be identified. For example, the first write data may be stored in a memory internal or external to the error type detection logic 130 and the first write data may be read from the memory for comparison with the second read data. According to embodiments described below, the memory controller 100 may identify the type of error and a stuck value.
A temporary error may be caused by drift of distribution according to accumulation of memory operations, disturb between a read operation and a write operation, and the like. For example, when the memory controller 100 performs an operation (for example, reclaiming) of rewriting the same data to a memory cell included in the memory device 200, the temporary error may be resolved. As another example, a temporary error may indicate that a data bit with an error is output in a first reading process, but a normal bit is output in a second reading process performed after the first reading process.
The memory controller 100 may pattern data HDTA received from the host, based on information about the identified error type and information about a memory area where the error has occurred.
According to an embodiment, the data patterning logic 140 may randomize the received data HDTA based on a random seed. The randomized data may have a data bit in each of a plurality of memory areas. In this case, certain memory areas may be memory areas where stuck errors have occurred. Each of the memory areas where the stuck errors have occurred may have a predetermined stuck value. Therefore, when, among data bits included in the data randomized by the data patterning logic 140, data bits included in the memory area where the stuck error has occurred have a high degree of agreement with the stuck value, it may be seen that the error is reduced. That is, the data patterning logic 140 may reduce the error by selecting data having the highest degree of agreement with the stuck value from among the randomized data.
Then, the data patterning logic 140 may output selected randomized data to the memory device 200. In this case, the data patterning logic 140 may add parity for decoding the randomized data thereto.
Referring to
According to the present embodiment, each of the plurality of memory cells MC may include a variable resistance device R and a switching device SW. The variable resistance device R may be referred to as a variable resistance material, and the switching device SW may be referred to as a selection device.
In an embodiment, the variable resistance device R may be connected between one of the plurality of bit lines BL1 to BLm and the switching device SW, and the switching device SW may be connected between the variable resistance device R and one of the plurality of word lines WL1 to WLn. However, the disclosure is not limited thereto, and the switching device SW may be connected between one of the plurality of bit lines BL1 to BLm and the variable resistance device R and the variable resistance device R may be connected between the switching device SW and one of the plurality of word lines WL1 to WLn.
The switching device SW may be connected between one of the plurality of word lines WL1 to WLn and the variable resistance device R and may control supply of electric current to the variable resistance device R according to a voltage applied to the connected word line and the bit line. In
Although
Referring to
The upper and lower electrodes 12 and 13 may include various metals, metal oxides, or metal nitrides. The upper and lower electrodes 12 and 13 may include aluminum (Al), copper (Cu), titanium nitride (TiN), titanium aluminum nitride (TixAlyNz), iridium (Ir), platinum (Pt), silver (Au), polysilicon, tungsten (W), titanium (Ti), tantalum (Ta), tantalum nitride (TaN), tungsten nitride (WN), nickel (Ni), cobalt (Co), chrome (Cr), antimony (Sb), iron (Fe), molybdenum (Mo), palladium (Pd), tin (Sn), zirconium (Zr), zinc (Zn), iridium oxide (IrO2), strontium oxide zirconate (StZrO3), or the like.
The phase-change layer 11 may include a bipolar resistance memory material or a unipolar resistance memory material. The bipolar resistance memory material is programmed to a set state or a reset state according to a polarity of an applied electric current and may include Perovskite-based materials. The unipolar resistance memory material is programmed to a set state or a reset state by an electric current having the same polarity and may include transition metal oxide such as NiOx or TiOx.
The GST material may be programmed between an amorphous state having a relatively high resistivity and a crystalline state having a relatively low resistivity. The GST material may be programmed by heating the GST material. The magnitude and time of the heating may determine whether the GST material remains in an amorphous state or a crystalline state. The high resistivity and the low resistivity may be represented by programmed values, that is, logic 0 and logic 1, respectively, and may be detected by measuring the resistivity of the GST material. Conversely, the high resistivity and the low resistivity may be represented by programmed values, that is, logic 1 and logic 0, respectively.
In
The memory controller 100 may read first read data RD1 from the memory device 200. For example, the first read data RD1 may be data written to at least one of the first memory area 210_1 to the Nth memory area 210_N. The ECC circuit 120 may perform error detection and error correction operations on the first read data RD1. For example, the first read data RD1 may be a codeword in which data bits and parity bits are combined, and as another example, the first read data RD1 may include data bits. The first read data RD1 may be outputted based on a read command.
According to an embodiment of the disclosure, when error correction fails, the ECC circuit 120 may perform a read retry operation by entering defensive code. Furthermore, when the error correction is successful, the ECC circuit 120 may write first write data WD1 to the memory device 200 when the number of error bits exceeds a threshold value. When the error correction is successful, the ECC circuit 120 may not write the first write data WD1 to the memory device 200 when the number of error bits is less than or equal to the threshold value. In this case, the ECC circuit 120 may determine that data is normally written to the memory device 200 to thereby keep the first read data RD1 in the memory device 200.
The ECC circuit 120 may perform error detection and error correction operations on the first read data RD1 and output error-corrected correction data CD. In addition, the ECC circuit 120 may identify a memory area (or the position of a memory area) where an error has occurred in a memory operation for the memory device 200 and output error bit information EBI. That is, the error bit information EBI may include various types of information which the ECC circuit 120 may derive as a result of performing error detection and error correction. The error bit information EBI may be included in stuck bit information. For example, the error bit information EBI may include information about a memory area (or the position of a memory area) where an error has occurred in the first read data RD1, information about an error value, and the like.
The error type detection logic 130 may output the first write data WD1 in response to receiving the correction data CD and the error bit information EBI. Specifically, the error type detection logic 130 may generate first write data WD1 that is inverted data of data (data included in the correction data CD) of another memory area other than an error-corrected memory area from among a plurality of memory areas. For example, the first read data RD1 may be [0, 1, 1, 0, 1] composed of 5 bits and the correction data CD may be [1, 0, 1, 0, 1] in which two bits of error are corrected. The first write data WD1 may be [1, 0, 0, 1, 0] in which the last three bits on which an error correction operation has not been performed are inverted. The above-mentioned bits and the number of bits are for convenience of explanation and the disclosure is not limited thereto. The error type detection logic 130 generates stuck bit information, which includes error bit information EBI and error type information, based on a result of performing a plurality of read and/or write operations of the memory device 200.
The memory controller 100 may write the first write data WD1 to the memory device 200. Then, error types may be detected by comparing second read data RD2 (see
Referring to
The error type detection logic 130 may perform an operation of comparing the second read data RD2 to the first write data WD1 and identify the error type of the second read data RD2 based on the error bit information EBI. For example, the error type detection logic 130 may identify an error type for each data bit included in the second read data RD2. As another example, by identifying the error type for each data bit included in the second read data RD2, the error type of a memory area corresponding to a data bit where an error has occurred may be identified.
The error type detection logic 130 may output error type information ETI, which is information about the identified error type. The error type information ETI may include information about the second read data RD2 and an error type related to at least one memory area included in the memory device 200. For example, the error type information ETI may include information about the type of error that occurred in each memory cell storing a data bit. The error type information ETI may be included in stuck bit information. The error type may include a stuck error, a temporary error, and a hidden stuck error, and each memory cell may be a normal cell or may have one of the error types. A method of identifying the error types will be described later with reference to
The data patterning logic 140 may randomize data HDTA received from the host, based on the error type information ETI. For example, the data HDTA may be new data to be written to the memory device 200. The data patterning logic 140 may select second write data WD2, which includes the least error, of randomized data and write the second write data WD2 to the memory device 200. According to an embodiment, a parity PA for decoding the randomized data may be generated. For example, the data patterning logic 140 may add the parity PA to the second write data WD2 and transmit the second write data WD2 with the parity PA to the memory device 200. Details will be described later with reference to
According to an embodiment of the disclosure, original data OD written to the memory device 200 may be [1, 0, 1, 0, 1]. The memory controller 100 may read the first read data RD1 based on a read request or an error detection request from the host. For example, the first read data RD1 may be [0, 1, 1, 0, 1].
An error may occur in data included in a first memory area 31 of the first read data RD1, and an error may not occur in data included in a second memory area 32 of the first read data RD1. For example, the first memory area 31 may be an area including first and second data bits, and the second memory area 32 may be an area including remaining data bits. In this case, the ECC circuit 120 may detect that an error has occurred in data included in the first memory area 31 and perform error correction by using a parity bit included in or outside the first read data RD1.
The ECC circuit 120 may perform error detection and error correction on the first read data RD1 and output the correction data CD. Specifically, the ECC circuit 120 may correct an error included in the first memory area 31 and maintain data included in the second memory area 32. For example, the correction data CD may be [1, 0, 1, 0, 1] and may be the same as the original data OD.
The error type detection logic 130 may generate the first write data WD1. Specifically, the error type detection logic 130 may maintain data of the first memory area 31, which is error-corrected in the correction data CD. In addition, the error type detection logic 130 may invert data of the second memory area 32, which is not error-corrected in the correction data CD. For example, the data bits [1, 0, 1], which are not error-corrected in the correction data CD, may be inverted to [0, 1, 0] and the inversion data bits may be included as part of the first write data WD1. The error type detection logic 130 may generate the first write data WD1 to perform an operation of comparing the first write data WD1 with the second write data WD2.
The error type detection logic 130 may output the first write data WD1 to be written to the memory device 200. Then, the memory controller 100 may read the second read data RD2 based on a request received from the host or a signal generated in the memory controller 100. The memory controller 100 may identify a position where the first write data WD1 is written to the memory device 200 and read the second read data RD2 that is data contained in the position. For example, the memory controller 100 may read the second read data RD2 based on the address ADD of the memory device 200 at which the first write data WD1 is written. Specifically, the error type detection logic 130 may compare data bits located at the same address ADD to each other.
The error type detection logic 130 may perform a comparison operation of comparing the first write data WD1 with the second read data RD2. According to an embodiment, the comparison operation may include a first partial comparison operation and a second partial comparison operation. The first partial comparison operation may be an operation of comparing the data of the first memory area 31 of the first write data WD1 to the data of the first memory area 31 of the second read data RD2. The second partial comparison operation may be an operation of comparing the data of the second memory area 32 of the first write data WD1 to the data of the second memory area 32 of the second read data RD2.
According to an embodiment of the disclosure, the error type detection logic 130 may compare data bits included in the first memory area 31 of the first write data WD1 to data bits included in the first memory area 31 of the second read data RD2. Referring to
The error type detection logic 130 may identify that a stuck error has occurred in a memory cell corresponding to a first partial area 41. Specifically, this is because, in the first partial area 41, the data bit [0] of the first read data RD1 and the data bit [1] of the original data OD are different from each other and the data bit [0] of the second read data RD2 read again after the first read data RD1 is read is also different from the data bit [1] of the first write data WD1. Each of a plurality of partial areas, that is, first to fifth partial areas 41 to 45, may correspond to one memory cell, or may correspond to a plurality of memory cells. An operation of comparing the first write data WD1 with the second read data RD2 may be referred to as a comparison operation, and an operation of comparing a partial area (e.g., the first partial area 41) of the first write data WD1 with a partial area, which is identical to the partial area of the first write data WD1, of the second read data RD2 may be referred to as a partial comparison operation.
The error type detection logic 130 may identify that a temporary error has occurred in a memory cell corresponding to the second partial area 42. Specifically, this is because, in the second partial area 42, the data bit [1] of the first read data RD1 and the data bit [0] of the original data OD are different from each other but the data bit [0] of the second read data RD2 read again after the first read data RD1 is read is the same as the data bit [0] of the first write data WD1. That is, the temporary error may be an error that occurs only temporarily in a process of reading the first read data RD1.
The error type detection logic 130 may identify that memory cells corresponding to the third partial area 43 and the fourth partial area 44 are normal. This is because no error has occurred in any of the first read data RD1 and the second read data RD2.
The error type detection logic 130 may identify that a stuck error has occurred in a memory cell corresponding to the fifth partial area 45. For example, the memory cell of the fifth partial region 45 may be identified as a memory cell in which a hidden stuck error, in which a data bit is stuck to 1, has occurred. The hidden stuck error may be a kind of stuck error. In the fifth partial area 45, the data bit of the original data OD may be [1] and the data bit of the first read data RD1 obtained by reading the original data OD written in the memory device 200 may also be [1]. However, in the fifth partial area 45, even if the data bit of the first write data WD1 is [0], the data bit of the second read data RD2 obtained by reading the first write data WD1 written in the memory device 200 may be [1]. That is, the error type detection logic 130 generates the first write data WD1 by inverting some data of the correction data CD, to identify the hidden stuck error.
According to embodiments of the disclosure, the memory controller 100 may identify whether the memory device 200 has an error, may identify a plurality of error types including a stuck error and a temporary error, and may also identify a hidden stuck error that is difficult to detect in a first reading process. As the memory controller 100 identifies various error types, the memory controller 100 may perform an operation of reducing errors occurring in the memory device 200 based on the identified error types. Accordingly, data reliability may be improved in a short time with only a single additional read operation and a simple comparison operation.
The memory controller 100 may: (1) write the second write data WD2 to the same memory cells of the memory device 200 from which it reads the second read data RD2; (2) read the second read data RD2 from the same memory cells of the memory device 200 in which it writes the first write data WD1; (3) write the first write data WD1 to the same memory cells of the memory device 200 from which it reads the first read data RD1; and (4) read the first read data RD1 from the same memory cells of the memory device 200 in which it writes the original data OD.
According to an embodiment of the disclosure, the error type detection logic 130 may identify, as a cell having a stuck error, a memory cell corresponding to an erroneous data bit in a first read operation and a second read operation. The error type detection logic 130 may determine a stuck value of at least one memory cell in which data identified as a stuck error is stored. For example, the stuck value may be a value of a data bit at which a stuck error has occurred in the second read data RD2.
According to an embodiment of the disclosure, when a data bit at which an error has occurred in the first read operation does not have an error in the second read operation, the error type detection logic 130 may identify a memory cell including the data bit as a cell having a temporary error. In addition, the error type detection logic 130 may identify, as a normal cell, a memory cell corresponding to a data bit at which no error has occurred in the first read operation and the second read operation.
According to an embodiment of the disclosure, when a data bit at which no error has occurred in the first read operation has an error in the second read operation, the error type detection logic 130 may identify a memory cell including the data bit as a cell having a hidden stuck error. For example, the stuck value may be a value of a data bit at which a hidden stuck error has occurred in the second read data RD2.
In an example, the error type detection logic 130 may store code for performing a series of operations to identify the error type and the stuck value. According to an embodiment, a memory (e.g., SRAM) may be included in or outside the error type detection logic 130 and the code stored in the memory may be read to perform the series of operations. As another example, the error type detection logic 130 itself may be software implemented with at least one piece of code. As another example, the error type detection logic 130 may perform a series of operations in a form (e.g., firmware) in which hardware and software are combined with each other. As another example, the error type detection logic 130 may identify the error type and the stuck value based on a look-up table that is the same as or similar to that shown in
Referring to
The random seed generator 141 may generate a plurality of random seeds used for randomizing data HDTA received from the host. For example, the random seeds may be fixed values or values that vary over time and may be values generated by random number generation. Furthermore, each of the plurality of random seeds may be applied to the received data HDTA, and a plurality of pieces of random data RDs may be generated by the number of random seeds. For example, as shown in
The randomizer 142 may randomize the received data HDTA based on the random seeds. The randomizer 142 may output the random data RDs to the minimum error selector 143. For example, the randomizer 142 may randomize the received data HDTA [1, 1, 0, 0, 1, 0, 1, 0] based on different random seeds and output different random data RDs. Data bits corresponding to first to eight partial areas 51 to 58 may be different between the different pieces of random data RDs. For example, the randomizer 142 may generate first random data having data of [1, 0, 0, 1, 1, 0, 1, 0] based on a first random seed and may generate second random data having data of [1, 1, 0, 1, 0, 1, 0, 1], which is different from the first random data, based on a second random seed. Each of the first to eight partial areas 51 to 58 may correspond to one memory cell, or may correspond to a plurality of memory cells.
The minimum error selector 143 may generate second write data WD2 based on the error type information ETI and the random data RDs. According to an embodiment, the minimum error selector 143 may generate parity PA for derandomizing (or decoding) the second write data WD2.
The minimum error selector 143 may select random data having the smallest error from among the pieces of random data RDs. For example, the minimum error selector 143 may select random data with the smallest stuck error, or may select random data with no stuck error. For example, the minimum error selector 143 may select sixth random data generated based on a sixth random seed from among the pieces of random data RDs. In the sixth random data, the first partial area 51, the third partial area 53, the fourth partial area 54, and the seventh partial area 57 have values 0, 1, 1, and 0, respectively. This is because memory cells corresponding to the first, third, fourth, and seventh partial areas 51, 53, 54, and 57 are stuck to have the stuck values of 0, 1, 1, and 0, and thus, the sixth random data may be data with no error. Therefore, the minimum error selector 143 may select the sixth random data.
For the sake of understanding, in the opposite case, when the memory controller 100 writes third random data generated based on a third random seed to the memory device 200, data to be written to the memory device 200 may be [0, 0, 1, 0, 1, 0, 1, 0]. However, data to be read from the memory device 200 may be [0, 0, 1, 1, 1, 0, 1, 0]. That is, a stuck error may occur in the data to be read due to a set stuck memory cell corresponding to the fourth partial area 54. Therefore, the minimum error selector 143 may select the sixth random data instead of the third random data.
According to an embodiment of the disclosure, the minimum error selector 143 may identify the address of memory cells into which the data of the first to eight partial areas 51 to 58 included in each of the pieces of random data RDs are to be written. Alternatively, the minimum error selector 143 may identify the address of memory cells into which the data HDTA received from the host is to be written. The minimum error selector 143 may determine whether a memory cell corresponding to the identified address is in an error state. In this case, the minimum error selector 143 may determine whether the memory cell is in an error state, based on the error type information ETI received from the error type detection logic 130. For example, the error type information ETI may include error types of memory cells, the position of a memory cell where an error has occurred, an error value, a stuck value, and the like. The minimum error selector 143 may refer to a stuck value of a memory cell corresponding to each of the first to eight partial areas 51 to 58 in the pieces of random data RDs and select random data in which a partial area corresponding to a memory cell with a stuck error has a data bit that matches the stuck value.
According to an embodiment of the disclosure, the minimum error selector 143 may output random data selected as the second write data WD2. The memory controller 100 may write the second write data WD2 to the memory device 200. For example, when the sixth random data is selected as the second write data WD2, the parity PA may be generated based on the sixth random seed. The generated parity PA may be used to derandomize (or decode) the second write data WD2 when the host wants to read the second write data WD2.
According to an embodiment of the disclosure, the data patterning logic 130 randomizes data HDTA. The data patterning logic 130 selects a first randomized data among a plurality of a random data RDs based on the error type information ETI. Data bits of the first randomized data include more data bits corresponding to a stuck bit than data bits of the first write data WD1.
According to an embodiment of the disclosure, the memory controller 100 may read data from the memory device 200 (S610). In this case, the memory controller 100 may read the first read data RD1, and at least some of the first read data RD1 may be different from the original data RD written in the memory device 200, due to various types of errors.
The memory controller 100 may perform an ECC operation on the read data (S620). For example, the ECC circuit 120 may perform error detection and error correction operations on the first read data RD1. As another example, an error of the first read data RD1 may be detected and corrected by an ECC engine included in the outside of the memory controller 100.
When the number of error bits included in the data read from the memory device 200 exceeds a threshold value, the memory controller 100 may invert data bits of the remaining area except for an error occurrence area and write the inverted data bits into the memory device 200 (S630). For example, referring to
Referring back to
The memory controller 100 may perform a randomizing operation (S650). For example, the memory controller 100 may randomize input data received from the host and may select random data, in which data to be written to the memory device 200 has a minimum error, based on the identified error. Then, the memory controller 100 may write the selected random data to the memory device 200 (S660). Operations S610 to S640 will be described later with reference to
Referring to
According to an embodiment of the disclosure, the memory controller 100 may output error bit information EBI about an error included in the first read data RD1 by performing error detection and error correction operations. For example, the error bit information EBI may include a position of a memory cell where an error has occurred, a position of a data bit at which an error has occurred in a data string, an error value, and an error-corrected correction value.
According to an embodiment of the disclosure, the memory controller 100 may invert remaining data bits except for corrected data bits (S631). For example, based on the error bit information EBI, the first write data WD1 may be generated by inverting remaining data bits except for some corrected data bits in the correction data CD.
According to an embodiment of the disclosure, the memory controller 100 may write data including the inverted data bits to the memory device 200 (S632). For example, the memory controller 100 may output the first write data WD1 to be written to the memory device 200. Some of the data bits of the first write data WD1 may include inverted data bits in the correction data CD and the remaining data bits of the first write data WD1 may be the same as the bits of the correction data CD.
According to an embodiment of the disclosure, the memory controller 100 may read the written data (S641). For example, the memory controller 100 may store information (for example, address) about the position of the memory device 200 in which the first write data WD1 is written, and may transmit, to the memory device 200, a read command for reading data of the position in which the first write data WD1 is written.
According to an embodiment of the disclosure, the memory controller 100 may perform a comparison operation of comparing data written to the memory device 200 with data read from the memory device 200. For example, the memory controller 100 may compare the first write data WD1 with the second read data RD2 to identify error types.
According to an embodiment of the disclosure, the memory controller 100 may identify whether a target for identifying an error type is a memory area of corrected data in the first write data WD1 and the second read data RD2 (S642). For example, the memory controller 100 may identify a memory area (e.g., the first memory area 31 in
When an error occurs again in the data of a memory area in which an error is corrected (S643), the memory controller 100 may identify, as a stuck error, the bit of an area in which the error has occurred (S645a). When an error does not occur in the data of a memory area in which an error is corrected, the memory controller 100 may identify, as a temporary error, the bit of an area in which no error has occurred (S645b). This is because the error is an error occurring only in a first reading process.
When an error occurs again in the data of a memory area in which an error is not corrected (S644), the memory controller 100 may identify, as a hidden stuck error, the bit of an area in which an error has occurred (S645c). This is because the error is an error occurring only in a second reading process. When an error does not occur in the data of a memory area in which an error is not corrected, the memory controller 100 may identify, as a normal bit, the bit of an area in which an error has not occurred (S645d).
The memory controller 100 may output information about an error associated with the memory device 200 and data read therefrom, obtained through a series of operations. That is, the memory controller 100 may output error type information (ETI) (S646).
Referring to
According to an embodiment of the disclosure, the memory controller 100 may select data having the smallest error from among the randomized data, based on the error type information (ETI) (S653). For example, when at least some data bits included in the pieces of random data RD are written to a memory cell in which a stuck error has occurred, the memory controller 100 may select random data in which the at least some data bits match a stuck value of the memory cell.
According to an embodiment of the disclosure, the memory controller 100 may write the selected data and parity corresponding thereto to the memory device 200 (S661). For example, the memory controller 100 may write the second write data WD2 to the memory device 200 and the parity PA may be data bits for derandomizing (or decoding) the second write data WD2.
Referring to
The processor 710 may include a central processing unit or a micro-processor and may control overall operations of the memory controller 700. For example, the processor 710 may be configured to drive software or firmware for controlling the memory controller 700 and the software or the firmware may be loaded on the RAM 720 to be operated. The RAM 720 may be used as an operation memory, a cache memory, or a buffer memory of the processor 710. The RAM 720 may temporarily store data that is to be written in the memory device or data read from the memory device.
The host interface 730 interfaces with the host to receive a request for memory operation from the host. For example, the host interface 730 receives various requests such as a read request, a write request of data, etc. from the host, and may generate various internal signals for performing memory operations on the memory device in response to the requests.
The memory interface 740 may provide an interface between the memory controller 700 and the memory device, for example, write data and read data may be transmitted to and received from the memory device via the memory interface 740. Also, the memory interface 740 may provide the memory device with commands and addresses and may receive various information from the memory device to provide the information to the inside of the memory controller 700.
The ECC circuit 750 may detect an error of read data through an ECC decoding process, and a result of detecting the error may be used in the operation of determining an error type in the error type detection logic 760. The data patterning logic 770 may generate a plurality of pieces of random data obtained by randomizing data received via the host interface 730. Based on information about an error type determined in the error type detection logic 760, the data patterning logic 770 may select random data with minimal error from among the pieces of random data. The selected random data may be output to the memory device 200 via the memory interface 740.
Various functions that may be performed by the error type detection logic 760 and the data patterning logic 770 may be implemented via hardware or software or a combination of the hardware and software. As an example, when a function is implemented via hardware, the function may be performed by a hardware circuit arranged in the inside of each of the error type detection logic 760 and the data patterning logic 770. Alternatively, when a function is implemented via software, a program provided in the inside or outside of each of the error type detection logic 760 and the data patterning logic 770 may be loaded on the RAM 720 and the processor 710 may execute the program to perform the function.
As is traditional in the field, embodiments may be described and illustrated in terms of blocks which carry out a described function or functions. These blocks, which may be referred to herein as units or modules or the like, are physically implemented by analog and/or digital circuits such as logic gates, integrated circuits, microprocessors, microcontrollers, memory circuits, passive electronic components, active electronic components, optical components, hardwired circuits and the like, and may optionally be driven by firmware and/or software. The circuits may, for example, be embodied in one or more semiconductor chips, or on substrate supports such as printed circuit boards and the like. The circuits constituting a block may be implemented by dedicated hardware, or by a processor (e.g., one or more programmed microprocessors and associated circuitry), or by a combination of dedicated hardware to perform some functions of the block and a processor to perform other functions of the block. Each block of the embodiments may be physically separated into two or more interacting and discrete blocks without departing from the scope of the disclosure. Likewise, the blocks of the embodiments may be physically combined into more complex blocks without departing from the scope of the disclosure. An aspect of an embodiment may be achieved through instructions stored within a non-transitory storage medium and executed by a processor.
While the disclosure has been particularly shown and described with reference to embodiments thereof, it will be understood that various changes in form and details may be made therein without departing from the spirit and scope of the following claims.
Kim, YoungSik, Lee, Jeongho, Lee, Younggeun, Yoo, Youngkwang, Baek, Seungyou, Oh, Eunchu
Patent | Priority | Assignee | Title |
Patent | Priority | Assignee | Title |
8054678, | Mar 17 2009 | Seagate Technology LLC | Stuck-at defect condition repair for a non-volatile memory cell |
8972649, | Oct 05 2012 | Microsoft Technology Licensing, LLC | Writing memory blocks using codewords |
8977929, | Feb 27 2013 | U S BANK NATIONAL ASSOCIATION, AS COLLATERAL AGENT | Rearranging write data to avoid hard errors |
9274884, | Oct 10 2012 | SanDisk Technologies, Inc | Encoding and decoding data to accommodate memory cells having stuck-at faults |
9582354, | Jan 28 2014 | Infineon Technologies AG | Apparatus and method for improving data storage by data inversion |
20120239996, | |||
20140157065, | |||
20140204671, | |||
20140347923, | |||
20150193304, | |||
20160098316, | |||
20160342467, | |||
20170271002, | |||
20180011754, | |||
20180210669, | |||
20180293131, | |||
20180343020, | |||
20190196896, | |||
JP5923631, | |||
KR1020180087496, |
Executed on | Assignor | Assignee | Conveyance | Frame | Reel | Doc |
Sep 20 2019 | YOO, YOUNGKWANG | SAMSUNG ELECTRONICS CO , LTD | ASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS | 052317 | /0362 | |
Oct 24 2019 | OH, EUNCHU | SAMSUNG ELECTRONICS CO , LTD | ASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS | 052317 | /0362 | |
Oct 31 2019 | LEE, JEONGHO | SAMSUNG ELECTRONICS CO , LTD | ASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS | 052317 | /0362 | |
Nov 10 2019 | BAEK, SEUNGYOU | SAMSUNG ELECTRONICS CO , LTD | ASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS | 052317 | /0362 | |
Nov 26 2019 | LEE, YOUNGGEUN | SAMSUNG ELECTRONICS CO , LTD | ASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS | 052317 | /0362 | |
Mar 30 2020 | KIM, YOUNGSIK | SAMSUNG ELECTRONICS CO , LTD | ASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS | 052317 | /0362 | |
Apr 06 2020 | Samsung Electronics Co., Ltd. | (assignment on the face of the patent) | / |
Date | Maintenance Fee Events |
Apr 06 2020 | BIG: Entity status set to Undiscounted (note the period is included in the code). |
Date | Maintenance Schedule |
Oct 12 2024 | 4 years fee payment window open |
Apr 12 2025 | 6 months grace period start (w surcharge) |
Oct 12 2025 | patent expiry (for year 4) |
Oct 12 2027 | 2 years to revive unintentionally abandoned end. (for year 4) |
Oct 12 2028 | 8 years fee payment window open |
Apr 12 2029 | 6 months grace period start (w surcharge) |
Oct 12 2029 | patent expiry (for year 8) |
Oct 12 2031 | 2 years to revive unintentionally abandoned end. (for year 8) |
Oct 12 2032 | 12 years fee payment window open |
Apr 12 2033 | 6 months grace period start (w surcharge) |
Oct 12 2033 | patent expiry (for year 12) |
Oct 12 2035 | 2 years to revive unintentionally abandoned end. (for year 12) |